Integrated Wastewater Treatment Plant Performance Evaluation Using Artificial Neural Networks

Abstract

Proper operation of municipal wastewater treatment plants is important in producing an effluent which meets quality requirements of regulatory agencies and in minimizing detrimental effects on the environment. This paper examined plant dynamics and modeling techniques with emphasis placed on the digital computing technology of Artificial Neural Networks (ANN). A backpropagation model was developed to model the municipal wastewater treatment plant at Ardiya, Kuwait City, Kuwait. Results obtained prove that Neural Networks present a versatile tool in modeling full-scale operational wastewater treatment plants and provide an alternative methodology for predicting the performance of treatment plants. The overall suspended solids (TSS) and organic pollutants (BOD) removal efficiencies achieved at Ardiya plant over a period of 16 months were 94.6 and 97.3 percent, respectively. Plant performance was adequately predicted using the backpropagation ANN model. The correlation coefficients between the predicted and actual effluent data using the best model was 0.72 for TSS compared to 0.74 for BOD. The best ANN structure does not necessarily mean the most number of hidden layers.

Simulation of an industrial wastewater treatment plant using artificial neural networks and principal components analysis

“…In the process industry the use of modern control strategies is required due to increasingly stringent regulation of effluent quality (Cohen et al, 1997;Lee and Park, 1999). Operational control of a biological wastewater treatment plant is often complicated because of variation in raw wastewater compositions, strengths and flow rates owing to the changing and complex nature of the treatment process (Hamoda et al, 1999). Moreover, a lack of suitable process variables limits the effective control of effluent quality (Harremoës et al, 1993, Lee andPark, 1999).…”

“…The modeling traditionally used in bioprocesses is based on balance equations together with rate equations for microbial growth, substratum consumption and formation of products, and since microbial reactions coupled with environmental interactions are nonlinear, time-variable and of a complex nature (Hamoda et al, 1999;Lee and Park, 1999), traditional deterministic and empirical modeling has shown some limitations (Cote et al, 1995;Hamoda et al, 1999).…”

“…Recently, some studies using artificial neural networks (ANNs) in modeling biological wastewater treatment processes have been published, providing an alternative approach (Cote et al, 1995;Häck and Kohne, 1996;Gontarski et al, 2000;Hamoda et al, 1999;Lee and Park, 1999;Pu and Hung, 1995;Wilcox et al, 1995;Zhao et al, 1997).…”

“…For modeling a wastewater treatment plant using neural networks, Hamoda et al (1999) found a correlation index of 0.74 for BOD prediction, Belanche et al (1999) found 0.504 for COD prediction and Häck and Köhne (1996) found 0.92 and 0.82 for COD and nitrate prediction, respectively.…”

Application of steady-state and dynamic modeling for the prediction of the BOD of an aerated lagoon at a pulp and paper millPart I. Linear approaches

“…Not surprisingly, the modeling traditionally used for bioprocesses, based on mass balance equations together with rate equations for microbial growth, substratum consumption and formation of products, has been shown to be inefficient for describing these mechanisms in wastewater treatment processes [3][4][5][6]. Therefore, in these cases when models based on first principles are not available, or requires excessive computation time, empirical models become attractive alternatives [7].…”